This repo contrains Jupyter notebooks that prepare and analyze data Canadian federal contracts over $10,000, which are published here (full data download here).
This analysis was used in two reports by the Investigative Journalism Foundation:
- Why do government IT contracts have such huge cost overruns?
- Ottawa spending record amounts outsourcing access to information requests
data_prep.ipynb - Loads raw data from government website, cleans up vendor names, standardizes unique procurement IDs, isolates amended contracts.
amended_contracts_analysis.ipynb - Statistical analysis of amended contracts and cost inflation.
Much of the prep code was inspired by the work of Sean Boots at Carleton University. For the most part, my Python code is a direct translation of Sean's R code.